基于深度学习的耕地变化检测技术
|
摘要
随着人工智能技术的发展,以深度学习为主流的机器学习逐渐取代人工解译的方法,使遥感影像中地物资源的自动化判读成为现实。为解决传统人工判读引起的人力资源耗费高、解析精度差的问题,同时也为满足日益增长的遥感数据量的判读需求,基于语义分割的深度学习地物变化检测方法,实现耕地区域自动分割分类,通过对比时序影像差异得出变化区域范围,为自动化实现地物变化监测提供有效解决办法。以实际地区为例,采用deeplab语义分割网络的方法实现耕地资源的自动化提取与变化检测,实验证明该方法相比人工以及传统分类模型具有更好的检测精度。
With the development of artificial intelligence technology,machine learning which takes deep learning as the mainstream has gradually replaced the manual interpretation method,making the automatic interpretation of terrestrial resources in remote sensing images become a reality.In order to solve the problems of high human resource consumption and poor resolution caused by traditional manual interpretation,and to meet the increasing demand for interpretation of remote sensing image data,the method of deep learning of landform change detection based on semantic segmentation is studied,the automatic segmentation and classification of cultivated land area are realized,the range of change area is obtained by comparing the differences of time series images,and the effective solution for automatically implementing landform change monitoring is provided.Taking Gaocheng District of Shijiazhuang as an example,the method of deeplab semantic segmentation network is used to realize automatic extraction and change detection of cultivated land resources.The experiment results show that this method has better detection accuracy compared with manual and traditional classification models.
With the development of artificial intelligence technology,machine learning which takes deep learning as the mainstream has gradually replaced the manual interpretation method,making the automatic interpretation of terrestrial resources in remote sensing images become a reality.In order to solve the problems of high human resource consumption and poor resolution caused by traditional manual interpretation,and to meet the increasing demand for interpretation of remote sensing image data,the method of deep learning of landform change detection based on semantic segmentation is studied,the automatic segmentation and classification of cultivated land area are realized,the range of change area is obtained by comparing the differences of time series images,and the effective solution for automatically implementing landform change monitoring is provided.Taking Gaocheng District of Shijiazhuang as an example,the method of deeplab semantic segmentation network is used to realize automatic extraction and change detection of cultivated land resources.The experiment results show that this method has better detection accuracy compared with manual and traditional classification models.
引文
[1] 赵忠明,孟瑜,岳安志,等.遥感时间序列影像变化检测研究进展[J].遥感学报,2016,20(5):1110-1125.
[2] 佟国峰,李勇,丁伟利,等.遥感影像变化检测算法综述[J].中国图象图形学报,2015,20(12):1561-1571.
[3] 黄春波,佃袁勇,周志翔,等.基于时间序列统计特性的森林变化监测[J].遥感学报,2015,19(4):657-668.
[4] 赵少华,刘思含,毛学军,等.新时期我国环保领域卫星遥感技术的应用与发展[J].无线电工程,2017,47(3):1-7.
[5] 颜洁,刘建坡,唐伟广,等.基于遥感图像变化检测的毁伤效果分析[J].无线电工程,2010,40(4):30-41.
[6] DOLLAR P,APPEL R,BELONGIE S,et al.Fast Feature Pyramids for Object Detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(8):1532-1545.
[7] 刘春华.遥感影像的半自动配准[J].无线电工程,2008,38 (5):32-34.
[8] 何梦梦.基于边缘点特征的高分辨率遥感图像配准方法研究[D].北京:中国科学院遥感与数字地球研究所,2017.
[9] 佃袁勇,方圣辉,姚崇怀.多尺度分割的高分辨率遥感影像变化检测[J].遥感学报.2016,20(1):129-137.
[10] 王月香,陈梅梅,冷燕艳.基于遥感的淮安市耕地变化分析[J].湖北农业科学,2013,52(21):5189-5192.
[11] 胡潭高,朱文泉,阳小琼,等.高分辨率遥感图像耕地地块提取方法研究[J].光谱学与光谱分析,2009,29(10):2703-2707.
[12] LONG J,SHELHAMER E,DARRELL T.Fully Convolutional Networks for Semantic Segmentation[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2014,39(4):640-651.
[13] GIRSHICK R,DONAHUE J,DARRELL T,et al.Region-Based Convolutional Networks for Accurate Object Detection and Segmentation[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2016,38(1):142-158.
[14] 陈志雄.基于图像配准的SIFT算法研究与实现[D].武汉:武汉理工大学,2008.
[15] 杨晓丽.多时相遥感影像的变化检测[D].西安:西安电子科技大学,2011.
[2] 佟国峰,李勇,丁伟利,等.遥感影像变化检测算法综述[J].中国图象图形学报,2015,20(12):1561-1571.
[3] 黄春波,佃袁勇,周志翔,等.基于时间序列统计特性的森林变化监测[J].遥感学报,2015,19(4):657-668.
[4] 赵少华,刘思含,毛学军,等.新时期我国环保领域卫星遥感技术的应用与发展[J].无线电工程,2017,47(3):1-7.
[5] 颜洁,刘建坡,唐伟广,等.基于遥感图像变化检测的毁伤效果分析[J].无线电工程,2010,40(4):30-41.
[6] DOLLAR P,APPEL R,BELONGIE S,et al.Fast Feature Pyramids for Object Detection[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2014,36(8):1532-1545.
[7] 刘春华.遥感影像的半自动配准[J].无线电工程,2008,38 (5):32-34.
[8] 何梦梦.基于边缘点特征的高分辨率遥感图像配准方法研究[D].北京:中国科学院遥感与数字地球研究所,2017.
[9] 佃袁勇,方圣辉,姚崇怀.多尺度分割的高分辨率遥感影像变化检测[J].遥感学报.2016,20(1):129-137.
[10] 王月香,陈梅梅,冷燕艳.基于遥感的淮安市耕地变化分析[J].湖北农业科学,2013,52(21):5189-5192.
[11] 胡潭高,朱文泉,阳小琼,等.高分辨率遥感图像耕地地块提取方法研究[J].光谱学与光谱分析,2009,29(10):2703-2707.
[12] LONG J,SHELHAMER E,DARRELL T.Fully Convolutional Networks for Semantic Segmentation[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2014,39(4):640-651.
[13] GIRSHICK R,DONAHUE J,DARRELL T,et al.Region-Based Convolutional Networks for Accurate Object Detection and Segmentation[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2016,38(1):142-158.
[14] 陈志雄.基于图像配准的SIFT算法研究与实现[D].武汉:武汉理工大学,2008.
[15] 杨晓丽.多时相遥感影像的变化检测[D].西安:西安电子科技大学,2011.